Question 1194542: Determine the type of sampling to the given scenario:
1.In Internal Revenue Service researcher investigates cheating on income tax reports by surveying all waiters and waitresses at 20 randomly selected restaurants. Which sampling type is this?
2.Motivated by a student who died from binge drinking, the College of Newport conducts a study of student drinking by randomly selecting 10 different classes and interviewing all of the students in each of thoseclasses. Which sampling type is this?
3.A market researcher has partitioned all California residents into categories of unemployed, employed full time, and employed part time. She is surveying 50 people from each category. Which sampling
type is ?
Answer by math_tutor2020(3816)   (Show Source):
You can put this solution on YOUR website!
Answers:
- Cluster sampling
- Cluster sampling
- Stratified sampling
Explanations:
Using the cluster sampling technique means we do the following steps- 1) Split the population into disjoint groups known as clusters. This means the groups do not overlap in any way, and there is no member left out. In short: there are no gaps and no overlaps
- 2) Randomly select as many clusters as you want. Use a random number table or software to properly do the random process as best as possible. Repeat selections are not allowed.
- 3) Once the clusters are selected, you will survey every person inside said clusters.
Usually clusters are geographic in nature such as states in the US. Though we could easily apply this to things like restaurants or classrooms (for problems 1 and 2)
Stratified sampling will also have us form disjoint groups. It may seem like this is similar to cluster sampling, and in some ways it is (which could explain why many students confuse the two methods). However, instead of randomly picking n groups to sample everyone from, we instead pick a small list of people from every subgroup. This ensures that every aspect of the population is covered. In the case of problem 3, stratified sampling ensures that people from the unemployed, employed full time, and employed part time groups are represented fairly.
Here's are examples illustrating the difference of cluster vs stratified sampling.
- Cluster: Pick 10 states at random, and survey everyone from those states chosen.
- Stratified: Survey 20 people from each of the 50 states (total sample size = 20*50 = 1000), which ensures every state in the US is represented fairly.
As you can probably tell, cluster sampling can get very expensive if the number of clusters gets too large and/or each cluster is large. A way to mitigate this is to try to reduce the size of each cluster, or simply select fewer clusters.
|
|
|
